Half of South African businesses see the value in data analysis to increase profit while more than a third rates cost reductions as a key benefit. But they have yet to fully realise the potential of analytics, says DESAN NAIDOO.
This was the key message from an analytics roundtable hosted by business analytics software and services provider SAS last week. Chaired by media personality Jeremy Maggs, panel members included Desan Naidoo, MD of SAS Southern Africa: Azar Jammine, chief economist at Econometrix: Professor Machiel Kruger, head of the SAS Advanced Analytics Lab at North West University: and Brent Henegen, BI manager: direct marketing and segmentation at Edcon, who discussed the findings of ‚ÄòThe State of Business Analytics in South Africa’ survey.
Conducted by SAS and BMI-TechKnowledge, the survey interviewed 350 company representatives to understand the state of business analytics in South Africa. Findings were compared to those of a similar survey conducted three years ago.
Naidoo noted that business analytics is at a premature stage in South Africa. Businesses use reactive tools such as spreadsheets (87%) and business dashboards (83%), which are easy to access and use. However, he said spreadsheets are “backwards-looking” and do not provide predictions that have a direct impact on competitiveness and productivity.
South African businesses have progressed in their application of analytics, said Naidoo. Compared to three years ago, local businesses are adopting a more integrated approach, moving from a siloed method that focuses on individual departments to applying it across the entire organisation (43%). Some 53% of respondents state that analytics contributes to more than 60% of their organisation’s decision-making. Another 8% indicated that analytics contributes to 0% to 20% of their decision-making.
Kruger distinguished between business intelligence (BI) and business analytics, saying the former has to do with data preparation and reporting while the latter was about predictions and using data to help businesses make decisions in a data-driven way. Elaborating further, Henegen said BI lets businesses look at averages while analytics lets them drill down into specific business areas to identify problem areas.
The importance of data integration was a common theme throughout the discussion. Jammine said integration enables effective decision-making. However, Kruger noted that integration is sometimes difficult to achieve as data must be “clean” in order to do reliable analysis. He defined clean data as information that has no missing variables, is up to date and is available from one source providing a “single view of the truth”.
A key challenge to achieving data integration is in eliminating data siloes, the panel noted. Henegen explained that different departments within a business may have their own datasets. These may have been captured differently (e.g. ID number vs birth date) and therefore the datasets don’t “speak to each other”. It is crucial that departments work together and share data, he said, to establish an analytics culture within a company.
This comes up against certain obstacles, said Naidoo. An analytics culture must be driven from an executive level. There should also be a strong education focus that explains the concepts of analytics to staff in plain language.
Once data integration has been achieved, businesses can start applying analytics to scenario development – only practiced by 22% of respondents. Naidoo said many businesses don’t see the value of analytics and scenario planning and typically go for the “quick wins” – 84% use it in the finance department to uncover fraud, for example. Other popular areas are strategy/planning (76%) and marketing (61%). Less than half use it for product development and human resources. However, when benefits are seen in these areas, analytics trickles down into other departments, eventually expanding across the entire organisation.
Jammine referred to the 2008 global financial crisis as an example of scenario planning. Those businesses that manipulated their data to experiment with different outcomes – and how those outcomes would affect their operations – would have been better prepared to respond to crises and, therefore, survive the economic slump.
Ongoing strikes in the platinum sector have resulted in the first contraction in South Africa’s economy since 2009. Businesses should use analytics to determine how to respond in different scenarios, be it further contraction or economic recovery. Analytics provides insight in tough economic times to see what’s going wrong and to fix it, said Kruger.
However, scenario planning is only effective when used alongside a strong data model, and formulating these models requires skill, noted Kruger. These skills include a working knowledge of mathematics, statistics and other technical and business knowledge. Further, students must “speak the language of executives” in order to get their buy-in. They should also be exposed to enterprise software during their studies so that they can “hit the ground running”. Some 64% of respondents were confident that they had the right analytical talent in place.
Touching on social media analytics, the panel agreed that businesses are not yet tapping into this wealth of information – only 20% of respondents use social media analytics. While Naidoo noted that social media is another “quick win” area, Kruger said social media data should be integrated with other, more reliable data to help with predictions. Henegen said it was difficult to get an accurate picture of a customer using only social media data. He said this data should be considered alongside traditional data to be useful.
The panel agreed that social media analytics was important as it allowed businesses to manage sentiment, react to crises and change brand perceptions.
Ultimately, those businesses that are not yet using data analytics will be forced into doing so by the market, said Kruger. It’s best to be an early adopter – before the competition.
* Desan Naidoo, MD of SAS Southern Africa
* Follow Gadget on Twitter on @GadgetZA